Visualization of Group Related Differences in Histopatholgy Data
Project Scope |
---|
Histopathology data comprise a major component of toxicology studies and are typically presented in data tables (tabular format) in study reports. Can these data be graphically displayed in a visual format which provides a means for rapidly identifying and communicating treatment group differences, patterns in response, and/or changes in phenotypic expression ? Can cross-domain analyses (including organ weights, clinical pathology and histopathology data) be graphically displayed for analyses ? Can data from multiple studies be aggregated or graphically displayed for analysis ? How does incorporation of the SEND (Standard for Exchange of Nonclinical Data) standard enable this technology? To canvas opinion from toxicologists, pathologists, bioinformatics specialists and software designers on current methods of data visualization and how those might be improved or expanded to improve data analyses compared to currently used tabular (incidence) formats. To communicate points of discussion, ideas, case examples, etc. with key stake holders (the pharmaceutical industry, software developers/vendors, CROs, US FDA) via the Advance Hub, poster presentations and publications. |
Vision |
---|
To explore the use of various software applications for visually displaying histopathology data in novel ways to aid with data analysis and communication of findings, and to promote the use of the SEND standard for enabling this technology |
Prior Project Plans |
---|
The project completed its third year of activity (2017 to 2018). Project plans included the following. 1) Evaluate the use of the "sunburst" graphical display (termed HistoGraphic) and application for visualizing histopathology data from the IMI eTOX database 3) Explore use of the "sunburst" (HistoGraphic) application for simultaneously displaying microscopic findings and relevant biomarkers. As an example, initial efforts can focus on liver lesions and serum ALT, AST, ALKP and/or bilirubin changes 3) Collaborate with the Nonclinical Scripts working group for development and evaluation of novel software applications, such as HistoGraphic, for data visualization. The first two projects listed above are a component of this activity 4) Create mock SEND datasets of histopathology data for incorporation into various software applications to evaluate multi-study data displays |
Summary of Projects |
---|
After 3 years of activity, the project has reached a natural conclusion with development of HistoGraphic for graphically displaying histopathology data from multiple toxicology studies which were obtained from the European Union's IMI eTOX database. Deliverables from the project team included 3 posters presented at FDA PHUSE CSS meetings and one publication. It is the project team's intent to place HistoGraphic on a publicly assessed site for downloading. Following completion of this project, a new project has been endorsed by PHUSE for initiation in 2018: Data Visualization as an Enabler for Nonclinical Safety Signal Detection. This project will include a collaborative relationship with the EU's IMI eTRANSAFE initiative (Enhancing Translational Safety Assessment through Integrative Knowledge Management) eTRANSAFE. |
Milestones |
---|
March 2015 - March 2016 (completed) 1) Identify a panel of Toxicologists and Pathologists (2 weeks) April 2016 - March 2017 (completed) 1) Prepare first draft of manuscript (July 2016) - completed April 2017 - March 2018 (completed) 1) Obtain eTOX datasets and incorporate into HistoGraphic display tool (completed November 2017) |
Data Visualisation Examples |
---|
|
Tasks |
---|
Upcoming milestones for 2018 FDA PHUSE Computational Sciences Symposium, 4-6 March 2018 in Silver Spring, Maryland, USA:
|
Deliverables for 2015 |
---|
Deliverables for 2016 |
---|
Publication in Regulatory Toxicology and Pharmacology Poster presented at 2017 FDA PHUSE Computational Sciences Symposium |